Skip to content

FabNEPTUNE is a FabSim3 plugin for automated NEPTUNE (UKAEA’s fusion project)-based simulations

License

Notifications You must be signed in to change notification settings

kbronik2017/FabNEPTUNE

 
 

Repository files navigation

GitHub issues GitHub forks GitHub stars

FabNEPTUNE



FabNEPTUNE is a tool that allows for the easy execution of NEPTUNE project simulations (especially convection2d and convection3d) in local or remote machines.

For the full FabNEPTUNE documentation, please visit https://fabneptune.readthedocs.io

How to run a Convection2D and Convection3D (test) Jobs

These examples assume that you have been able to run the basic FabSim examples described in the other documentation files, and that you have built and configured Nektar++ (https://www.nektar.info/) on the target machine.

Two minimal examples of Convection2D and Convection3D simulation are provided in config_files/SCEMa_test1 and config_files/convection_2d_test and config_files/convection_3d_test to execute these examples type:

fabsim localhost Convection2D_local:convection_2d_test
fabsim <remote machine name> Convection2D_remote:convection_2d_test
fabsim localhost Convection3D_local:convection_3d_test
fabsim <remote machine name> Convection3D_remote:convection_3d_test

Run Ensemble Examples

Convection2D_ensemble

These examples assume that you have been able to run the basic FabSim examples described in the other documentation files, and that you have built and configured Nektar++ (https://www.nektar.info/) on the target machine.

To run type:

fabsim localhost Convection2D_ensemble_local:Convection2D_ensemble_example
fabsim <remote machine name> Convection2D_ensemble_remote:Convection2D_ensemble_example

This example runs 3 simulations with different input files, which vary the simulation timestep, using the same topology file.

Convection3D_ensemble

This example runs 3 simulations with different input files, which vary the simulation timestep, using the same topology file.

To run type:

fabsim localhost  Convection3D_ensemble_local:Convection3D_ensemble_example
fabsim <remote machine name>  Convection3D_ensemble_remote:Convection3D_ensemble_example

EasyVVUQ+FabNEPTUNE

After updating the following files with your credentials

  -FabSim3/deploy/machines_user.yml
  -FabSim3/deploy/machines.yml
  -FabSim3/plugins/FabNEPTUNE/machines_FabNEPTUNE_user.yml
  

run the following:

  -  fabsim   localhost   Convection2D_init_run_analyse_campaign_local:convection_2d_easyvvuq_InRuAn1_QCGPJ
  -  fabsim   <remote machine name>   Convection2D_init_run_analyse_campaign_remote:convection_2d_easyvvuq_InRuAn1_QCGPJ
  -  fabsim   localhost   Convection2D_init_run_analyse_campaign_local:convection_2d_easyvvuq_InRuAn2_QCGPJ
  -  fabsim   <remote machine name>   Convection2D_init_run_analyse_campaign_remote:convection_2d_easyvvuq_InRuAn2_QCGPJ

and copy the results back to your local machine with

 -  fabsim  localhost   fetch_results
 -  fabsim  <remote machine name>   fetch_results

EasyVVUQ+EasySurrogate+FabNEPTUNE

After updating the following files with your credentials

  -FabSim3/deploy/machines_user.yml
  -FabSim3/deploy/machines.yml
  -FabSim3/plugins/FabSCEMa/machines_FabNEPTUNE_user.yml
  

run the following:

  -  fabsim   localhost   Convection2D_init_run_analyse_campaign_local:convection_2d_easyvvuq_easysurrogate_InRuAn1_DAS_QCGPJ
  -  fabsim   <remote machine name>   Convection2D_init_run_analyse_campaign_remotel:convection_2d_easyvvuq_easysurrogate_InRuAn1_DAS_QCGPJ
  -  fabsim   localhost   Convection2D_init_run_analyse_campaign_local:convection_2d_easyvvuq_easysurrogate_InRuAn2_DAS_QCGPJ
  -  fabsim   <remote machine name>   Convection2D_init_run_analyse_campaign_remote:convection_2d_easyvvuq_easysurrogate_InRuAn2_DAS_QCGPJ

and as always copy the results back to your local machine with

 -  fabsim  localhost   fetch_results
 -  fabsim  <remote machine name>   fetch_results
 

About

FabNEPTUNE is a FabSim3 plugin for automated NEPTUNE (UKAEA’s fusion project)-based simulations

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • C++ 51.1%
  • Python 47.2%
  • Shell 1.7%